In this proposal we describe experiments to extend our previous analysis of the scratch (scrt) gene which encodes a pan-neural transcription factor likely to function as a repressor of non-neuronal gene expression (Specific Aim l). We also propose to follow a new line of investigation focused on the analysis of the short gastrulation (sog) gene, which we isolated in the course of our studies of pan-neural enhancer elements. sog is expressed in the early neuroectoderm and encodes likely antagonist of the TGFbeta-like Dpp signaling pathway (Specific Aims II-IV). Sog and the Dpp signaling pathway have been highly conserved during evolution. Dpp and its vertebrate homologue BMP-4 can functionally substitute for each other in dorsalventral pattern formation in flies and in bone morphogenesis in vertebrates. In addition, a Xenopus homologue of sog known as chordin has been identified recently. Like sog and dpp in flies, chordin and BMP-4 also have opposing effects on dorsal-ventral patternIng. Thus, it is likely that the sog,/chordin and dpp/BMP-4 genes play conserved roles in early dorsal-ventral patterning in invertebrates and vertebrates. These studies are therefore highly relevant to understanding the regulation of growth factors in the TGFbeta superfamily. Specific research goals are to: I. Determine how the pan-neural genes scratch and deadpan collaborate to promote neurogenesis. II. Determine whether sog is a dedicated inhibitor of Dpp signaling in flies and frogs. III. Identify enhancer element(s) directing sog expression during embryonic and adult development. IV. Determine whether a Sog protein(s) is secreted, diffuses, and binds to Dpp.